基于张量表示的高光谱图像目标检测

张小荣; 潘志斌; 郑茜

doi:10.3788/OPE.20192702.0488

您当前的位置：

首页 >

文章列表页 >

基于张量表示的高光谱图像目标检测

信息科学 | 更新时间：2020-08-13

- 基于张量表示的高光谱图像目标检测
- Tensor representation based target detection for hyperspectral imagery
- 光学精密工程 2019年27卷第2期页码：488-498
- 作者机构：
  
  1.中国科学院西安光学精密机械研究所，陕西西安 710119
  2.西安交通大学电子信息与工程学院，陕西西安 710049
  3.中国科学院大学，北京100049
  4.中国科学院地球环境研究所，陕西西安 710016
- 作者简介：
  
  [ "张小荣(1984-), 男, 陕西神木人, 博士研究生, 助理研究员, 2007、2010年于西北工业大学分别获得学士、硕士学位, 主要从事高光谱图像分类、目标检测方面的研究。E-mail:zhangxiaorong@opt.ac.cn" ]
- 基金信息：
  
  国家自然科学基金(61501456);陕西省自然科学基金(2018JM6065)
- DOI：10.3788/OPE.20192702.0488
  中图分类号： TP751
- 收稿日期：2018-07-14，
  
  录用日期：2018-9-19，
  
  纸质出版日期：2019-02-15
- 稿件说明：
移动端阅览
张小荣, 潘志斌, 郑茜. 基于张量表示的高光谱图像目标检测[J]. 光学精密工程, 2019,27(2):488-498.

Xiao-rong ZHANG, Zhi-bin PAN, Xi ZHENG. Tensor representation based target detection for hyperspectral imagery[J]. Optics and precision engineering, 2019, 27(2): 488-498.
张小荣, 潘志斌, 郑茜. 基于张量表示的高光谱图像目标检测[J]. 光学精密工程, 2019,27(2):488-498. DOI： 10.3788/OPE.20192702.0488.

Xiao-rong ZHANG, Zhi-bin PAN, Xi ZHENG. Tensor representation based target detection for hyperspectral imagery[J]. Optics and precision engineering, 2019, 27(2): 488-498. DOI： 10.3788/OPE.20192702.0488.

摘要

高光谱图像目标检测作为一个研究热点在军事和民用方面的应用越来越广泛。为了能同时利用高光谱图像数据的空谱信息，本文提出一种新的基于张量表示的高光谱图像目标检测算法。算法使用CP(Canonical Polyadic)张量分解技术和张量块分解(Block Term Decomposition，BTD)分别对高光谱数据进行盲源分析，提取了有效的局部图像块空谱特征，建立了一个基于稀疏表示和协作表示的检测模型，针对多种类型背景复杂的场景数据进行实验，并与当前流行的目标检测算法进行比较。从可视化检测结果来看，本文算法在复杂背景和强噪声环境下，有效提取了空谱特征，对背景具有较好的抑制能力，检测的目标显著。此外，本文从接收机操作曲线(Receiver Operating Characteristic Curve，ROC)和ROC曲线下面积(Area Under Curve，AUC)等定量指标分析算法性能。以较为流行的Sandiego图像为例，在10%的虚警率下，本文算法取得90%的检测精度，AUC大于0.95。本文算法相较几种流行算法而言具有较高的检测精度，更强的鲁棒性。

Abstract

Target detection for Hyperspectral Images (HSIs) is gaining importance owing to its important military and civilian applications. This study proposed a novel target detection algorithm for HSIs based on tensor representation. The algorithm employed tensor analysis including CP and tensor block decompositions to implement blind source separation on hyperspectral data. First

effective spatial and spectral features of the blocks of local images were extracted. Then

a detection model based on sparse and collaborative representations was established. Experiments were conducted to evaluate the performance of our approach under multiple scenes with complex backgrounds. From the visual representation of the results

it can be concluded that the proposed approach effectively extracts the spatial-spectral features from scenes with strong noise and complex backgrounds. The approach has good ability to suppress the background and the target is salient. In addition

the performance of the approach is evaluated using quantitative metrics such as Receiver Operating Curve (ROC) and area under the ROC curve (AUC). Considering the popular HSI image of San Diego as an example

the approach achieves 90% detection rate with a false alarm rate of 10%

and the AUC is greater than 0.95. Hence

our approach outperforms other popular approaches.

关键词

Keywords

references

唐意东, 黄树彩, 凌强, 等.高光谱图像自适应核联合表示异常检测[J].强激光与粒子束, 2015, 27(9): 49-55.

TANG Y D, HUANG SH C, LING Q, et al .. Adaptive kernel collaborative representation anomaly detection for hyperspectral imagery [J]. High Power Laser and Particle Beams , 2015, 27(9): 49-55. (in Chinese)

赵春晖, 靖晓昊, 李威.基于StOMP稀疏方法的高光谱图像目标检测[J].哈尔滨工程大学学报, 2015, 36(7): 992-996.

ZHAO CH H, JING X H, LI W. Hyperspectral image target detection algorithm based on StOMP sparse representation [J]. Journal of Harbin Engineering University , 2015, 36(7): 992-996. (in Chinese)

赵春晖, 孟美玲, 李威.基于稀疏表示的高光谱图像增殖快速算法[J].黑龙江大学自然科学学报, 2017, 34(1): 95-102.

ZHAO CH H, MENG M L, LI W. Hyperspectral imagery target detection proliferative fast algorithm based on sparse representation [J]. Journal of Natural Science of Heilongjiang University , 2017, 34(1): 95-102. (in Chinese)

凌强, 黄树彩, 韦道知, 等.联合表示求解二元假设模型的高光谱目标检测[J].电子学报, 2016, 44(11): 2633-2638.

LING Q, HUANG SH C, WEI D ZH, et al .. Collaborative representation-based binary hypothesis model for hyperspectral target detection [J]. Acta Electronica Sinica , 2016, 44(11): 2633-2638. (in Chinese)

BITAR AW, CHEONG LF, OVARLEZ JP. Sparse and low-rank decomposition for automatic target detection in hyperspectral imagery [J]. Electrical-Engineering and Systems Science , 2017, 24(11): arxiv: 1711.08970. https://www.researchgate.net/publication/321278542_Sparse_and_Low-Rank_Matrix_Decomposition_for_Automatic_Target_Detection_in_Hyperspectral_Imagery

NIUY, WANG B. Hyperspectral anomaly detection based on low-rank representation and learned dictionary [J]. Remote Sensing , 2016, 8 (4): 289.

Nasrabadi NM. Hyperspectral target detection: an overview of current and future challenges [J]. IEEE Signal Processing Magazine , 2014, 31 (1): 34-44.

AKBARI D, SAFARI A. Support vector machine for target detection in hyperspectral images [J]. TS06I-Remote Sensing II , 6135, 2012: 10.

黄鸿, 陈美利, 段宇乐, 等.空-谱协同流形重构误差的高光谱影像分类[J].光学精密工程, 2018, 26(7): 1827-1836.

HUANG H, CHEN M L, DUAN Y L, et al .. Hyper-spectral image classification using spatial-spectral manifold reconstruction [J]. Opt. Precision Eng ., 2018, 26(7): 1827-1836. (in Chinese)

MAKANTASISK, KARANTZALOS K, DOULAMIS A, et al . Deep learning-based man-made object detection from hyperspectral data [C]. Cham: Springer International Publishing , 2015: 717-727. https://www.researchgate.net/publication/289193167_Deep_Learning-Based_Man-Made_Object_Detection_from_Hyperspectral_Data

ZHANG L, ZHANG L, TAO D, et al .. Hyperspectral remote sensing image subpixel target detection based on supervised metric learning [J]. IEEE Transactions on Geoscience and Remote Sensing , 2014, 52(8): 4955-4965.

Transter learning [EB/OL]. http://en.m.wikipedia.org/wiki/Transfer_learning http://en.m.wikipedia.org/wiki/Transfer_learning .

DU B, ZHANG L, TAO D, et al .. Unsupervised transfer learning for target detection from hyperspectral images [J]. Neurocomputing , 2013, 120: 72-82.

DONG Y, DU B, ZHANG L, et al .. Local decision maximum margin metric learning for hyperspectral target detection [C]. 2015 IEEE International Geoscience and Remote Sensing Symposium, Milan, Italy: IGARSS , 2015: 397-400. https://ieeexplore.ieee.org/document/7325784/

REN Y, LIAO L, MAYBANK S J, et al .. Hyperspectral image spectral-spatial feature extraction via tensor principal component analysis [J]. IEEE Geoscience and Remote Sensing Letters , 2017, 14(9): 1431-1435.

谷延锋, 高国明, 郑贺, 等.高分辨率航空遥感高光谱图像稀疏张量目标检测[J].测绘通报, 2015(1): 31-38.

GU Y F, GAO G M, ZHENG H, et al .. High solution airborne hyperspectral remote sensing images target detection via tensor and sparse [J]. Bulletin of Surveying and Mapping , 2015(1): 31-38. (in Chinese)

ZHANG Q, WANG H, PLEMMONS RJ, et al . Tensor methods for hyperspectral data analysis: a space object material identification study [J] . Journal of the Optical Society of America A , 2008, 25(12): 3001-3012.

VEGANZONESMA, COHEN JE, FARIAS RC, et al .. Nonnegative tensor CP decomposition of hyperspectral data [J]. IEEE Transactions on Geoscience and Remote Sensing , 2016, 54(5): 2577-2588.

LIU Y, GAO G, GU Y. Tensor matched subspace detector for hyperspectral target detection [J]. IEEE Transactions on Geoscience and Remote Sensing , 2017, 55(4): 1967-1974.

ZHANG X, WEN G, DAI W. Target representation in hyperspectral images based on tensor block term decomposition [C]. 2015 8th International Congress on Image and Signal Processing, Shenyang, China: CISP , 2015: 793-798. https://www.researchgate.net/publication/304409624_Target_representation_in_hyperspectral_images_based_on_tensor_block_term_decomposition

LATHAUWERL D, NION D. Decompositions of a higher-order tensor in block terms—part Ⅲ: alternating least squares algorithms [J]. SIAM Journal on Matrix Analysis and Applications , 2008, 30(3): 1067-1083.

LI W, DU Q, ZHANG B. Combined sparse and collaborative representation for hyperspectral target detection [J]. Pattern Recognition , 2015, 48 (12): 3904-3916.

BIOUCAS-DIAS JM, FIGUEIREDO MAT. Alternating direction algorithms for constrained sparse regression: Application to hyperspectral unmixing [C]. 2010 2nd Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing, Reykjavik, Iceland : 2010: 1-4. http://www.oalib.com/paper/3916880

LI W, DU Q. Collaborative representation for hyperspectral anomaly detection [J]. IEEE Transactions on Geoscience and Remote Sensing , 2015, 53(3): 1463-1474.

QIAN D, HSUAN R, CHEIN I C. A comparative study for orthogonal subspace projection and constrained energy minimization[J]. IEEE Transactions on Geoscience and Remote Sensing , 2003, 41 (6): 1525-1529.

ZOUZ, SHI Z. Hierarchical suppression method for hyperspectral target detection [J]. IEEE Transactions on Geoscience and Remote Sensing , 2016, 54 (1): 330-342.

BACHER R, MEILLIER C, CHATELAIN F, et al .. Robust control of varying weak hyperspectral target detection with sparse nonnegative representation [J]. IEEE Transactions on Signal Processing , 2017, 65 (13): 3538-3550.

WANG Y, PENG J, ZHAO Q, et al .. Hyperspectral Image restoration via total variation regularized low-rank tensor decomposition [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2018, 11 (4): 1227-1243.

YUKSELS E, KARAKAYA A. Fusion of target detection algorithms in hyperspectral images [J]. International Journal of Intelligent Systems and Applications in Engineering , 2016, 4 (4): 103-110.

YANG S, SHI Z. Hyperspectral image target detection improvement based on total variation [J]. IEEE Transactions on Image Processing , 2016, 25 (5): 2249-2258.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于张量表示的高光谱图像目标检测算法

跨层注意力交互下的多特征交叉无人机图像检测

适应大范围星场密度变化的恒星辨识与抑制

利用光谱解混合的目标检测

基于分数阶微分的高光谱图像特征提取与分类